Machine learning / 2026 / Runs natively
Resume Lens
A resume analyser that reads a CV the way an applicant tracking system does, scores each bullet on Google's X-Y-Z formula, and measures coverage of a pasted job description with BM25 term weights.
01
The problem
Resume scanners reject people without ever saying why, and simple keyword checkers credit "experienced" as an Experience section.
02
How I approached it
Sections count only under a heading line a parser would route, keywords match spelling variants but flag where the exact term is missing, bullets earn a third each for action verb, measurement and method, and job-description terms are weighted on the BM25 saturation curve with a boost for named skills.
03
The outcome
A score where every point traces back to a named rule, plus the job terms the CV never uses.
Related work
A 26 KB neural network running entirely in a browser tab. Draw a digit, it classifies on device, and the page reports which execution provider it actually got and what the inference cost.
Always-on desktop assistant listening for a wake phrase, then routing the spoken task straight into Gemini.
Captures whatever is on screen, asks Gemini to solve it, then types the answer into the app you're focused on.